94
GitHub - hengruo/QANet-pytorch
source link: https://github.com/hengruo/QANet-pytorch
Go to the source link to view the article. You can view the picture content, updated content and better typesetting reading experience. If the link is broken, please click the button below to view the snapshot at that time.
README.md
QANet-pytorch
Introduction
An implementation of QANet with PyTorch. Now it can reach EM/F1 = 70.5/77.2 after 20 epoches for about 20 hours on one 1080Ti card.
Usage
Python 3.6 & PyTorch 0.4
- Install pytorch 0.4 for Python 3.6+
- Run
pip install spacy tqdm ujson requests - Run
python -m spacy download en - Run
python main.py
Structure
dataset.py: download dataset and parse.
main.py: program entry.
models.py: QANet structure.
Differences from the paper
- The paper doesn't mention which activation function they used. I use relu.
- I don't set the embedding of
<UNK>trainable. - The connector between embedding layers and embedding encoders may be different from the implementation of Google, since the description in the paper is inconsistent (residual block can't be used because the dimensions of input and output are different) and they don't say how they implement it.
TODO
- Performance analysis
- Reach state-of-art scroes of the original paper
- Ablation analysis
- Improvement
Recommend
About Joyk
Aggregate valuable and interesting links.
Joyk means Joy of geeK